About Me


Hi! My name is Karan. I am a Data Scientist with ~4 years of experience in the field of Data Science and Machine Learning. My academic background is in the field of Computer Science with a specialization in Data Science. I am a graduate student of NC State University with a Master's degree in Computer Science.

I am interested in solving business problems through data-driven solutions. Data is a mystery and I love solving this mystery. I love data wrangling, data visualization, and building machine learning models to solve business problems.

I love to play basketball and Table Tennis. I also like traveling to new places and exploring new cuisines.


Education and Experience



  Download My Résume

    View and Download


   Programming

Python
R
Pandas
Numpy
Dplyr
Javascript
Tensorflow

  Data Engineering

SQL
PostgreSQL
MongoDB
Apache Spark
Apache Kafka

  Cloud technologies & Other tools

AWS RDS
AWS Redshift
Azure
Azure Data Factory
Azure Data Studio
Azure Machine Learning
Docker
Github Actions

  Tools

Tableau
Power BI
Github
Azure

  Certifications

Machine Learning by IBM
Data Analysis by IBM
Deep Learning specialization
Big Data with Pyspark

  • Quadient Inc

    Data Scientist

    • Led a churn prediction project leveraging ML models (Logistic Regression, Random Forest, AdaBoost) achieving an F1 score of 0.83 and a recall of 0.79. We analyzed individual segments to determine strategies to retain high risk customers.
    • Marketing Cluster Analysis: Helped the marketing team to prioritize important partners for the products using K-Means clustering algorithm. It helped to save $200K/year by segmenting the correct partners.
    • Designed and executed A/B testing to evaluate the effectiveness of email marketing campaigns resulting in a 15% increase in conversion rate from baseline.
    • Built automated ETL pipelines using python scripts, scraping techniques, APIs to prepare data sources for reporting, dashboards and data analysis saving 10+ hours of manual work every week.
    • Identified cross sell opportunities for sales & marketing based on product usage and customer firmographics and bringing 14% new customers and revenue for business.
    • Built ~10 Dashboards & Reports to track KPI’s & identify trends for Postal & Mailing systems.

  • Greenlight Biosciences Inc.

    Data Scientist

    • Leading analytics projects for Data Science & research teams to optimize the production platform & DNA sequences.
    • Built an automated risk assessment pipeline that helped the production team to predict & forecast the performance variables for gene sequences.
    • Analyzed the data using Tableau and build an analytics dashboard for external stakeholders.
    • Built a fast & reliable CI-CD pipeline to automate deployment of applications saving deployment time by 20%.

  • Greenlight Biosciences Inc.

    Data Science Intern

    • Created DNA performance metrics dashboard that involved performing ETL on 200K+ data points using SQL and R.
    • Optimized and contributed new features which increased the overall accuracy by 20% and reduced time by 50%.
    • Streamlined the data pipelines by deploying Shiny applications using Docker to ease access to reports & visualizations.
    • Conducted a bioinformatics workshop to present non-technical groups the features of applications and dashboards.

  • North Carolina State University, Raleigh

    Masters Degree

    Masters in Computer Science with courses relevant to Data Science & Machine Learning.

  • NMIMS University

    Bachelors of Engineering

    I completed my Bachelors in Computer Engineering from NMIMS University.


Projects & Blogs


Check some of my interesting projects!

  • All
  • Data Science & ML
  • NLP
  • Recommender Systems
Pier 1

Email Sender Classification


The project focuses on identifying the sender of the email based on the email context.
Pier 1

Hashtag Propounder


This project focuses on building a hashtag recommender system to suggest hashtags based on images and captions through Object Detection and Natural Language Processing.
Pier 1

Stock Insights


StockInsight predicts the stock prices using Neural Networks. It uses Elasticsearch, Twitter data, News Headlines and sentiment analysis to find the effect of emotions on the stock price.
Pier 1

Terrain Identification


This project focuses on identifying the terrain surface based on the gyroscope and accelerometer sensors.
Pier 1

Movie Recommendation


A Movie recommender system based on content based and collaborative based algorithms. The recommender system is deployed on AWS using Flask.
Pier 1

Music Recommendation


The project focuses on creating a music recommendation system that will suggest artists to user based on their listening history. The project uses a collaborative filtering technique and uses Apache Spark and Python for implementation.